To expand on what Allen is saying: It is very enlightening to experiment with VituixCad2 using ideal driver responses, and then simulate a 2-way speaker in a box.
Ideal drivers are the "default" option when you first open VituixCad, the drivers have a perfectly flat response. The next step is to use the diffraction tool and create some baffle shape, and then simulate the horizontal and vertical responses of the ideal tweeter and ideal woofer.
The next step is to build a crossover filter using the active elements available in VituixCad, such as low pass, high pass, notch, shelf, etc. Experimenting with active elements eliminates the complexities of impedance and electrical phase, and it allows us to concentrate on just the acoustical behavior of the simulation.
We might expect that with ideal, perfectly flat drivers, it would be very simple to get a flat response... however, if we simply slapped on a basic LR2, BW3, or LR4 crossover, we would not get a satisfactory result. We will often need quite a bit of filtering to achieve a good response. Baffle step is one obvious area we will have to deal with, but there are other diffraction effects as well. There are also acoustical interactions between the woofer and tweeter depending on their spacing, and on the crossover frequency and slope.
If the goal is a flat on-axis response and a good directivity index behavior, we might need quite a bit of experimenting in VituixCad to get a good simulation. What I found is that there are some baffle shapes/layouts which are easy to work with, and getting a good result is simple... There are other baffle shapes/layouts which are difficult to work with.
If there is an interest, I can post some of the more interesting results I have found... but I don't want to clog up this thread unless there is an interest.
j.
Ideal drivers are the "default" option when you first open VituixCad, the drivers have a perfectly flat response. The next step is to use the diffraction tool and create some baffle shape, and then simulate the horizontal and vertical responses of the ideal tweeter and ideal woofer.
The next step is to build a crossover filter using the active elements available in VituixCad, such as low pass, high pass, notch, shelf, etc. Experimenting with active elements eliminates the complexities of impedance and electrical phase, and it allows us to concentrate on just the acoustical behavior of the simulation.
We might expect that with ideal, perfectly flat drivers, it would be very simple to get a flat response... however, if we simply slapped on a basic LR2, BW3, or LR4 crossover, we would not get a satisfactory result. We will often need quite a bit of filtering to achieve a good response. Baffle step is one obvious area we will have to deal with, but there are other diffraction effects as well. There are also acoustical interactions between the woofer and tweeter depending on their spacing, and on the crossover frequency and slope.
If the goal is a flat on-axis response and a good directivity index behavior, we might need quite a bit of experimenting in VituixCad to get a good simulation. What I found is that there are some baffle shapes/layouts which are easy to work with, and getting a good result is simple... There are other baffle shapes/layouts which are difficult to work with.
If there is an interest, I can post some of the more interesting results I have found... but I don't want to clog up this thread unless there is an interest.
j.
^Yes! please do, people don't understand the importance of "the construct" until experimenting by them selves like you have. Even with ideal drivers and simple baffle model without the whole box and problems within and real world drivers and rooms, most of the "performance" is in the construct, how the system comes together. Add in the complexity and limitations of real world stuff and the importance just got bigger. The whole system needs to play together for best outcome. Suitable drivers and perfected xo is just part of it. Premium parts have effect but miniscule in comparison.
Maybe few examples would possibly perhaps motivate at least some to try them selves. Most would think it is too much work and just go by the easy stuff like bass box alignments, brand names and two part passive crossovers. Not to point fingers to anyone, many people just don't see the complete picture what makes loudspeaker System. For those of us who want to tinker and get the best sound it is kind of obvious
But it is a hobby for fun so everyone should go by their own comfort of course.
edit.
I've tried to post as many VCAD demos as possible about the construct and various phenomena, but is a lot of work to do and is scattered all around on relevant topics. And there are lot more here an there made by fluid, AllenB, nc535 and many others I don't happen to remember. Should there be a dedicated topic to collect this kind of demos under one hood for people to notice better? I think there is very much valuable insight available on the demos, void of marketing and mythology and would love to see people taking advantage of.
Maybe few examples would possibly perhaps motivate at least some to try them selves. Most would think it is too much work and just go by the easy stuff like bass box alignments, brand names and two part passive crossovers. Not to point fingers to anyone, many people just don't see the complete picture what makes loudspeaker System. For those of us who want to tinker and get the best sound it is kind of obvious
But it is a hobby for fun so everyone should go by their own comfort of course.edit.
I've tried to post as many VCAD demos as possible about the construct and various phenomena, but is a lot of work to do and is scattered all around on relevant topics. And there are lot more here an there made by fluid, AllenB, nc535 and many others I don't happen to remember. Should there be a dedicated topic to collect this kind of demos under one hood for people to notice better? I think there is very much valuable insight available on the demos, void of marketing and mythology and would love to see people taking advantage of.
Last edited:
Agreed. I try my best to measure thoughtfully, so that I can figure out exactly what is a correctable issue - and what is related to the box or the room
i am not sure i understood this correctly. Could this timing change potentially be compensated by eg rePhase?
Timing/phase mismacth related to changing distance (and frequency) of two sources cannot be fixed with rePhase (in signal before xo). Summed response of two separate radiators can be optimized with it at one fixed direction, but off-axis is always a compromise and not perfect IRL.
Still, direct sound from a loudspeaker (summed response of a multiway) is what dominates what we hear. Smooth off-axis/DI, low distortion and low EGD come after that. I haven't tested rePhase or a similar FIR, so I don't really know how much better it can make a poor or a good loudspeaker's sound.
Great will look into that one.
I was playing around a bit with vituixCad myself.
You mention to use diffration tool.
What I did so far for simulation and learning the program.
Trace IEC baffle responses.
Used the FRD and ZMA files in the diffraction tool to adapt them to the baffle.
I put the drivers on their X,Y on the baffle and the mic at tweeter height.
Then used that output in the enclosure tool to adapt them to enclosure volume.
and TADA my final FRD and ZMA files which I can use in my simulation.
Does this make sense or did i use the tool in the wrong way?
I think my wide baffle (40x100cm) is close to wide baffle.